Table A.1.
The tables from (a) to (e) contains the P-values of Nemenyi post-hoc test from all the comparisons across all the scanners, when evaluated by (a) single scanner training, (b) all scanners training, (c) inter-scanner normalization (Macenko), (d) inter-scanner fine tuning, and (e) HCT and HAC methods or experiments. The overall results with help of plots are presented in Fig. A.3 for a prompt overview.
| Trained on A | Trained on B | |||||||||
| A | B | C | D | A | B | C | D | |||
| A | 1 | 0.532 | 0.900 | 0.001 | A | 1 | 0.001 | 0.002 | 0.001 | |
| B | 0.532 | 1 | 0.199 | 0.001 | B | 0.001 | 1 | 0.900 | 0.001 | |
| C | 0.900 | 0.199 | 1 | 0.001 | C | 0.002 | 0.900 | 1 | 0.001 | |
| D | 0.001 | 0.001 | 0.001 | 1 | D | 0.001 | 0.001 | 0.001 | 1 | |
| Trained on C | Trained on D | |||||||||
| A | B | C | D | A | B | C | D | |||
| A | 1 | 0.001 | 0.001 | 0.001 | A | 1 | 0.001 | 0.001 | 0.001 | |
| B | 0.001 | 1 | 0.297 | 0.001 | B | 0.001 | 1 | 0.732 | 0.340 | |
| C | 0.001 | 0.297 | 1 | 0.001 | C | 0.001 | 0.732 | 1 | 0.039 | |
| D | 0.001 | 0.001 | 0.001 | 1 | D | 0.001 | 0.340 | 0.039 | 1 | |
| (a) Single scanner training | ||||||||||
| Trained on combined set (A, B, C, D) | ||||||||||
| A | B | C | D | |||||||
| A | 1 | 0.002 | 0.046 | 0.001 | ||||||
| B | 0.002 | 1 | 0.001 | 0.001 | ||||||
| C | 0.046 | 0.001 | 1 | 0.013 | ||||||
| D | 0.001 | 0.001 | 0.013 | 1 | ||||||
| (b) All scanners training | ||||||||||
| Trained and normalized on A | Trained and normalized on B | |||||||||
| A | B | C | D | A | B | C | D | |||
| A | 1 | 0.311 | 0.270 | 0.257 | A | 1 | 0.001 | 0.057 | 0.244 | |
| B | 0.311 | 1 | 0.900 | 0.900 | B | 0.001 | 1 | 0.609 | 0.244 | |
| C | 0.270 | 0.900 | 1 | 0.900 | C | 0.057 | 0.609 | 1 | 0.900 | |
| D | 0.257 | 0.900 | 0.900 | 1 | D | 0.244 | 0.244 | 0.900 | 1 | |
| Trained and normalized on C | Trained and normalized on D | |||||||||
| A | B | C | D | A | B | C | D | |||
| A | 1 | 0.001 | 0.001 | 0.008 | A | 1 | 0.001 | 0.001 | 0.001 | |
| B | 0.001 | 1 | 0.037 | 0.001 | B | 0.001 | 1 | 0.900 | 0.111 | |
| C | 0.001 | 0.037 | 1 | 0.025 | C | 0.001 | 0.900 | 1 | 0.244 | |
| D | 0.008 | 0.001 | 0.025 | 1 | D | 0.001 | 0.111 | 0.244 | 1 | |
| (c) Inter-scanner normalization | ||||||||||
| Trained on A and fine-tuned for each scanner | Trained on B and fine-tuned for each scanner | |||||||||
| A | B | C | D | A | B | C | D | |||
| A | 1 | 0.900 | 0.007 | 0.468 | A | 1 | 0.900 | 0.001 | 0.900 | |
| B | 0.900 | 1 | 0.053 | 0.150 | B | 0.900 | 1 | 0.001 | 0.900 | |
| C | 0.007 | 0.053 | 1 | 0.001 | C | 0.001 | 0.001 | 1 | 0.001 | |
| D | 0.468 | 0.150 | 0.001 | 1 | D | 0.900 | 0.900 | 0.001 | 1 | |
| Trained on C and fine-tuned for each scanner | Trained on D and fine-tuned for each scanner | |||||||||
| A | B | C | D | A | B | C | D | |||
| A | 1 | 0.001 | 0.001 | 0.900 | A | 1 | 0.283 | 0.900 | 0.001 | |
| B | 0.001 | 1 | 0.066 | 0.001 | B | 0.283 | 1 | 0.593 | 0.066 | |
| C | 0.001 | 0.066 | 1 | 0.001 | C | 0.900 | 0.593 | 1 | 0.001 | |
| D | 0.900 | 0.001 | 0.001 | 1 | D | 0.001 | 0.066 | 0.001 | 1 | |
| (d) Inter-scanner fine-tuning | ||||||||||
| HAC without normalization | HAC normalized to scanner A | |||||||||
| A | B | C | D | A | B | C | D | |||
| A | 1 | 0.563 | 0.002 | 0.233 | A | 1 | 0.092 | 0.001 | 0.870 | |
| B | 0.563 | 1 | 0.001 | 0.900 | B | 0.092 | 1 | 0.001 | 0.386 | |
| C | 0.002 | 0.001 | 1 | 0.001 | C | 0.001 | 0.001 | 1 | 0.001 | |
| D | 0.233 | 0.900 | 0.001 | 1 | D | 0.870 | 0.386 | 0.001 | 1 | |
| HCT without normalization | HCT with normalization to B | |||||||||
| A | B | C | D | A | B | C | D | |||
| A | 1 | 0.001 | 0.578 | 0.013 | A | 1 | 0.001 | 0.402 | 0.419 | |
| B | 0.001 | 1 | 0.001 | 0.043 | B | 0.001 | 1 | 0.001 | 0.001 | |
| C | 0.578 | 0.001 | 1 | 0.283 | C | 0.402 | 0.001 | 1 | 0.900 | |
| D | 0.013 | 0.043 | 0.283 | 1 | D | 0.419 | 0.001 | 0.900 | 1 | |
| (e) HCT and HAC methods | ||||||||||